The Agent Skills Directory

[SAFE]: The skill implements a deterministic policy evaluation engine. The primary functionality is contained within scripts/simulate_invocation.py, which processes command-line arguments to simulate routing decisions.
[SAFE]: All operations are performed locally using standard Python libraries (argparse, json, math, re, pathlib). The script does not use dangerous functions like eval() or exec(), nor does it perform any network requests or subprocess spawning for external command execution.
[SAFE]: File access is restricted to reading the policy configuration file provided as a command-line argument. The logic within scripts/simulate_invocation.py correctly handles JSON parsing and implements validation for policy fields (e.g., threshold ranges, boolean checks, and list types).
[SAFE]: The instructions and documentation correctly define the scope of the skill as an auditable activation policy router and explicitly state that activation loads instructions but does not approve tools or filesystem changes.

skill-invocation-router